Is Weight-Bearing Asymmetry Associated with Postural Instability after Stroke? A Systematic Review

Table 1

Study characteristics.

Author (year)	Study design	Number of participants with stroke	Participant characteristics; (1) Age (years); (2) Time after stroke	Baseline measures of stroke severity	Baseline measures of clinical balance; /ambulatory performance	Category of balance measure	Dependent measures of postural stability	Statistics for association between WBA and postural stability
Van Asseldonk et al. (2006) [29]	Cross-sectional	𝑁=8	(1) 59.9 (8.3); (2) 42.4 (17.1) months	MI 51.5 (16.1)	FAC (4-5); BBS 49 (7.6); TUG 21.8 (11.9); TBT (3–4)	DP	Contribution of paretic leg to; balance control^a	↑ 𝑅^2=0.26
Mansfield et al. (2011) [10]	Cross-sectional	𝑁=33	(1) 61 (14); (2) 133 (194) days	NIHSS 4.6 (4.1); ChMcS-leg 4.6 (1.3); ChMcS-foot 4.2 (1.6)	BBS 37.9 (15.6)	QS	COP amplitude synchronization^b	↑ 𝐑^𝟐  = 0.19–0.23
Mansfield et al. (2012) [30]	Cross-sectional; Retrospective chart; Review	𝑁=100	(1) 66.9 (14.9); (2) 22.9 (23.5) days	NIHSS 3.5 (3.0); ChMcS-leg 4.8 (1.3); ChMcS-foot 4.5 (1.4)	BBS 36.9 (14.6)	QS; CB; FA	COP amplitude synchronization^b; BBS; Faller status	↑ 𝑅^2=0.032–0.053; ↑ 𝐑^𝟐  = 0.049–0.057; Multiple linear regression; 𝑃=0.060; Multiple logistic regression
Marigold et al. (2004) [26]	Cross-sectional	𝑁=10	(1) 61.3 (8.9); (2) 4.1 (2.9) years	Not reported	BBS 44.9 (8.3)	DP	TA amplitude; MG amplitude; TA onset latencies; MG onset latencies	F2,18 = 1.51(p), 0.67(np); ↑ 𝐅𝟐,𝟏𝟖  =  𝟔,𝟏𝟒(𝐩),  𝟑,𝟕𝟖(𝐧𝐩); Anova; F2,18 = 0.15(p), 0.78(np); F2,18 = 0.19(p), 0.70(np)
Marigold and Eng (2006) [25]	Cross sectional	𝑁=28	(1) 62.1 (8.6); (2) 4.1 (2.7) years	ChMcS-leg 4 (3-5); ChMcS-foot 5 (4–6)	BBS 46.2; (42.4–48.6) IQR	QS	COP amplitude (AP); COP velocity (AP); COP amplitude (ML); COP velocity (ML)	↑ 𝑅^2=0.0049; ↑ 𝑅^2=0.01; ↑ 𝑅^2=0.044; ↑ 𝐑^𝟐  = 0.22
Pereira et al. (2010) [27]	Cross-sectional	𝑁=14	(1) 65 (10); (2) 29 (23) months	Not reported	Not reported	CB	Functional Reach	NS total group; ↓ 𝑅^2 = 0.10
Peurala et al. (2007) [28]	Cross-sectional	𝑁=45	(1) Left hemi 53.4 (8); Right hemi 55.1 (8); (2) Left hemi 3.1 (4.8) years; Right hemi 2.8 (2.8) years	FIM; Left hemi; 103.4 (10); Right hemi; 101.1 (13)	FAC (2–5)	QS	COP velocity (AP); COP velocity (ML); Power peak magnitude (AP)^c; Power peak magnitude (ML)^c	↑ 𝐑^𝟐  = 0.18,; ↑ 𝐑^𝟐  = 0.25; ↑ 𝐑^𝟐  = 0.09–0.14; ↑ 𝐑^𝟐  = 0.12–0.20
Roerdink et al. (2009) [24]	Cross-sectional*	𝑁=33	(1) 61.2 (13.0); (2) 9.8 (5.4) weeks	BFM; II–VI	FAC 1–4	QS	Sway area^d	↑ 𝑅^2=0.073
Sackley (1991) [8]	Longitudinal cohort	𝑁=92	(1) 63.3 (21–87); (2) 11.3 (5) weeks	Not reported	Not reported	FA	Number of falls	NS

Values for descriptive data represent mean (sd) unless stated otherwise. Statistics represent 𝑅² as calculated from bivariate correlations unless stated otherwise. Significant associations are presented in bold. ↑: greater WBA is associated with more postural instability, ↓: greater WBA is associated with less postural instability AP: anteroposterior, BBS: Berg Balance Scale, BFM: Brunnstrom Fugl-Meyer assessment, CB: clinical balance test, ChMcS: Chedoke-McMaster Stroke assessment scores, COP: center of pressure, DP: dynamic posturography, FA: falls in daily life, FAC: Functional Ambulation Categories, FIM: Functional Independence Measure, FR: Functional Reach test, IQR: interquartile range, MG: medial gastrocnemius, MI: Motricity Index, ML: mediolateral, N: number of stroke participants, NIHSS: National Institutes of Health Stroke Scale, NP: non-paretic, NS: no significant association with WBA, p: paretic, QS: quiet standing posturography, TA: tibialis anterior, TBT: Timed Balance Test, TUG: Timed Up and Go test, WBA: weight-bearing asymmetry. *Longitudinal study, but correlations between WBA and postural control were analyzed cross-sectionally. ^a% contribution of the paretic leg to total amount of generated corrective torque. ^bCalculated by cross-correlating COP amplitude time series, ^cpower spectral density functions, ^dtotal area covered by the COP trajectory.